The problem
This section reviews an anomaly detection problem using the NYC Taxi Traffic dataset from Kaggle (https://www.kaggle.com/datasets/julienjta/nyc-taxi-traffic), sourced from the NYC Taxi and Limousine Commission. This dataset contains univariate time series observations of the total number of taxi passengers between July 2014 and January 2015, aggregated at 30-minute intervals. The data include five anomalies during the NYC Marathon, Thanksgiving, Christmas, New Year’s Day, and a snowstorm.
You will implement end-to-end anomaly detection by analyzing the NYC Taxi Traffic dataset, creating a Long Short Term Memory (LSTM) model to predict outliers, and explaining anomalies using an OmniXAI SHAP explainer.
The following section reviews a step-by-step walk-through for this example.